Understanding How to Train

Download PDF  7–10 minutes to read

There are many different ways of training dogs, but these can all be split into 2 basic categories. This guide will help you understand each, and how each does or doesn’t work.

There are other ways of training, such as non-training, (used in assistance dog training) but these are very specialist and in depth training techniques and are generally based on positive reinforcement.

Dominance, pack or alpha dog style training

This school of training uses the idea that because dogs are descended from wolves, they obviously have to behave and have the same social structure as wolves. Therefore owners should ensure that at all times they are the ‘alpha’ in the relationship.

There are 2 main issues with this theory. In 2005 the dog genome was decoded, thanks to a boxer called Tasha. This means we know what every part of a dogs DNA looks like, but not what every little bit does. The vast majority of dogs’ DNA is the same as a wolf. Does this mean they are wolves? Well… humans share 50% of the same genes as a Banana plant, do you feel the need to hang out in a yellow suit, with a bunch of other people in yellow suits, upside down on a plant? Genes are tricky things, even the human genome is not fully understood, nor how each set of genes interacts with each other. The genetic evidence shows that dogs and wolves diverged on the evolutionary tree at the same time, one line ending up as the Grey Wolf, and the other as Fido.

The second issue is possibly the most important. The basis for this type of training is a study performed in the 1970’s in which unrelated Timber Wolves were forced to cohabitate and the results were observed, and extrapolated. The key word in this is unrelated, wolves live in familial groups, and even when a wolf leaves to find a mate and start its own pack the new pair have a long and very complex emotional bonding process. A typical wolf pack consists of parents, their siblings, and their children. The unrelated wolves in this study showed use of excessive dominance behaviours, and a constant desire for higher social standing. Do wolves really need to be in charge all the time? This behaviour is not seen in wild populations, nor even in captive family groups, so why did this group behave this way? Quite simply, they had to survive. With no relationships with the other wolves, each wolf became solely interested in his own survival, and each wolf forcibly maintained its position and the position of others by force and displays of dominance. These wolves were in a high stress environment, maintaining a dysfunctional group dynamic using threat and deference displays to keep safe, and get food.

Dr David Mech, the scientist largely responsible for this study has himself since renounced this study.  In a ‘true’ wolf pack, as in a family of dogs, submission is given to the parents and siblings freely, and is never forced.

So it’s kind of flawed, and most people who practice it will helpfully ignore the science and use anecdotal evidence to support their work. So how does it work, and does it work?

Dominance training requires humans to lead their dogs to believe they are the ‘top dog’ in the relationship. The person constructs a social hierarchy for the dog, and the dog is given a rank.

In this school it is considered dominant if a dog goes through a door first, pulls on the lead, doesn’t let you take its food away or guards its toys. If a dog is scared of other dogs, particular situations, or is overly excitable the human is deemed to not have enough dominance over the dog. Dominance is gained by a person, or reduced in a dog, by using corrections on unwanted or inappropriate behaviours, and by using intimidating body language towards the dog. Corrections include tugs on slip leads, use of prong or shock collars, a manoeuvre called a ‘bite’ where a sharp, fast jab to the throat or abdomen is applied using the fingertips, an alpha roll or a kick to the abdomen. Reward training is very seldom used by ‘pack leaders’, as it is seen as bribery. This type of training is often called “Balanced dog training”, implying that a well-balanced dog will be the result.

Do the corrections work? Well, to most people it would appear so, and they appear to work well and quickly. Recent studies have shown that dogs trained using these techniques are anything but balanced. The main idea of corrections is that you are interrupting a behaviour as or ideally before it begins, and the dog learns that behaviour is unacceptable. So, as an example, lets say your dog is scared of other dogs, and barks his head off as soon as he sees one. You don’t have enough dominance over the dog, and must use the corrections mentioned above to inform the dog that you are in control, and he has no need to worry about anything because you are in charge and will make all his decisions for him. You would be required to walk past strange dogs using lead corrections, abdomen kicks and possibly ‘biting’, and in extreme cases alpha rolls. Your dog would quickly quieten down when walking near strange dogs, and you would be chuffed with the trainer you paid a fortune for. Fast forward 2 months. You think, “hey my dog is perfect now, I think I’ll take him to meet another dog face to face” So off you go to the dog park, and your dog is totally silent when approaching the dog park, he stands by your side and makes no move to go towards the other dogs. A good sign, so you go in. Next thing you know you’re paying for 3 other dogs vet bills, and having your dog put down. Why? Because he is still scared. So scared he would risk serious physical harm at the jaws of another dog than tell you this. But he didn’t show any signs, why? Because of all those wonderful corrections. You told him that you don’t want to see him being defensive, don’t care that he’s scared out of his mind, you just don’t want him to warn you or other dogs that he’s not happy. And he listened, and paid with his life.

Corrections teach a dog to go into a state known as “Learned helplessness”, where a dog essentially shuts down all emotional responses that have been ‘corrected’, and becomes a placid ‘follower’.

Positive reinforcement, reward based or positive-only training

This type of training relies on the idea that dogs need rewards for doing as they are asked, or for making the correct decision in a given situation.

Rewards can be anything a dog goes mad for – be it food, a favourite toy or even a big cuddle.

Isn’t that just bribery? Won’t the unwanted behaviour return when the reward stops?

No, and no.

To understand why we have to look at a dogs brain architecture and chemistry. There are circuits on a dog’s brain that encourage seeking or hunting behaviours and circuits that encourage fear responses. The seeking circuits are triggered by and regulated by Dopamine. Dopamine is a neurotransmitter that is key in reward-based learning, and helps regulate movement and emotional responses. Dopamine also disengages the fear centres of the brain, even if only for a short time. In that short initial break, a dogs seeking circuits are engaged, and over time he learns that this behaviour (such as standing next to a terrifying car) makes him feel good, and the longer he practices, the better he feels and the less his fear controls him.

Using rewards for obedience training works in the same way as it does in behaviour modification. They learn that a particular cue means they get a dopamine rush, and they feel happy.

So what happens when the food stops, surely the dopamine is meaningless then?

This is covered quite in-depth in other guides, but in summary when a cue is reliably taught rewards can be reduced and the behaviour will still work. This process relies on the part of a dog’s brain that seeks. It’s the same part in humans that makes gambling addictive – we didn’t win this time, or the last 10 times, so we HAVE to hit the jackpot next time. This will encourage them to respond faster and more reliably, in the hopes of getting the jackpot. Eventually all rewards for a certain behaviour can be reduced to a ‘low value’ given fairly infrequently and each time the dog will get that happy feeling that dopamine gives him.

Some dominance type trainers refer to this school of training as ‘positive-only’ under the impression that there are no bad consequences for unwanted behaviours.

Note the word bad. Reinforcement training isn’t just about shoving treats in a dogs mouth and hoping for the best. Often it involves redirecting behaviours, or removing an item, or more powerfully, removing a person from a situation.

Redirecting is a great way of taking an unwanted behaviour and giving the dog the option to use the behaviour on something appropriate. If you have a dog who gets bossy and over excited when playing with other dogs, instead of pulling him away and shouting at him, offer him a tug toy instead. Then he can work out all his excitement on something appropriate and you both get a game of tug of war too!

Removing an item, attention or person is a negative consequence to unwanted behaviour. It is the fastest and most reliable way to stop puppy nipping for example. As soon as a puppy nips and the person leaves the room and shuts the door, the puppy stops and wonders what’s going on. After a few more nips and the puppy realises it doesn’t get to play with people if it bites them. A dog that jumps up can be taught not to by simply crossing your arms, turning your body away and not looking at it. Dogs crave attention, so this method very quickly imprints upon them that attention is given when they provide certain behaviours.

So which is best?

As with everything, no one can answer that but you. One thing trainers, especially fear based trainers, say is that there is no right way to train a dog. And that is true. But there are a million wrong ways to do it, and fear is one of them.

Science has proven that fear, intimidation and pain are not compatible with long term learning. Nor are they compatible with trust. Trust your dog, teach it to choose for itself, and it will trust you and do as much as it can to make you happy.

If you need more guidance of training methods please get in touch.